
Insight Integration for HP OpenView Network Node Manager version 3.1

This package contains a complete installation of the Insight Integration
for OpenView NNM version 3.1.  No other updates are required after 
running the installation.

-----

User Guide change: 

Step 10a on page 2-6 should read as follows.
a. Change to the %OV_WWW_REG%\dynamicViews\%LANG% directory.


---------------------------------------------------------------------

What's New in the integration:

 - Support for Network Node Manager version 7.x (versions 7.1 thru 7.5)
 - Added the cpqservice.mib.
 - Added separate icons for Microsoft Windows Server 2003
 - Added separate icons for Red Hat Linux and SuSE Linux
 - Added default support for HP Systems Insight Manager
 - The trap definition files have been updated to include support up 
   to the 7.10 version of the HP Insight Management Agents.
 - Updated support for discovery of HP Integrity Superdome systems.

 - Events are no longer deleted before install, so any customization to
   the 3.0 integration trap definitions will be preserved.

   Trap definitions are loaded with the xnmevents -load option.
   Any definition in file which already exists in trapd.conf will be 
   ignored. (Events are matched if they have the same name, identifier,
   and list of nodes.)  This option preserves all user customizations 
   to existing event definitions (in trapd.conf).

 
---------------------------------------------------------------------

Directory Contents:

The following directories and contents are provided with the 3.1
integration from HP:

bin - contains the updated binaries for the integration background
      and foreground processes.
      
bitmaps - contains the icons used by the integration.

config - contains the cpqconfig.dat file for discovery of HP systems
         and the CpqWeb.conf file for configuring which brower to use.

docs - contains the PDF user guide for the integration.

dynamicviews - contains the integration menu defintion used in the NNM
               dynamic views.

fields - contains the field registration files used by the integration.

hpqtraps - contains the files necessary to install a trap only version
           of the integration.  These files are not used in conjunction
           with the full integration.
           
ipf - contains the files necessary for supporting alarms from HP
      Integrity Superdome systems.

lrf - contains the registration files for the two background processes
      used by the integration (cpqdis and cpqtrapd).

mibs - contains the ProLiant MIBs.  By default, these MIBs are now 
       installed by the integration.       
       
registration - contains the application and menu registration information.
       
symbols - contains updates to the symbol registration files.

tools - contains tools that can be used to help verify the installation
        of the integration.

traps - contains updates to the trap definition files.  These files are
        used with the full integration.
        
www - contains GIFs and registration files for use with the java GUI.

---------------------------------------------------------------------

Installing:

Run "cpqinstall.cmd" from the integration directory. This will install/update
the Insight Integration.  Note: this will replace any existing Compaq trap 
definitions with new definitions. 


Uninstalling:

Remove the integration symbols from the maps using Map - Properties - Applications.
Run "cpquninstall.cmd" to remove the integration.  


-----

Configuration:


Systems Insight Manager TDEF file:

Edit the tool definition file nnm.xml and perform a search and replace
on the string "localhost".  Replace localhost with the name or IP address
of your Network Node Manager server.



Alarm Browser Views:

Edit the %OV_CONF%\%LANG%\xnmeventExt.conf file and add entries for web 
agent and Systems Insight Manager views.  These entries are shown next.
Replace "localhost" with the name or IP address of your Systems Insight
Manager server.

.1.3.6.1.4.1.232.*;"Insight Web Agents";http://$OvNode:2301

.1.3.6.1.4.1.232.*;"Systems Insight Manager";https://localhost:50000/mxportal/MxContextLaunch.jsp?systems=$OvNode&tool=DEFAULT

Any event from the 232 enterprise will now have the option to launch to the web
agents and Systems Insight Manager from the Actions -> Views menu.

Note: These entries will appear for every alarm from the 232 enterprise.



Dynamic Views Menu:

Edit the insight.xml file and perform a search and replace on the string
"localhost".  Replace localhost with the name or IP address of your
Systems Insight Manager server.


-----

Loading/unloading the MIBs:

By default, the MIBs are loaded when the integration is installed.
Other scripts are provided for loading and unloading the MIBs if
necessary.

Run "cpqloadmibs.cmd" from the \MIBS directory to load the mibs into NNM.
If you load the mibs with xnmloadmibs, do not load the trap definitions,
click cancel when prompted.  The trap definitions are provided in the 
translation files and do not need to be loaded from the mibs.

Run "cpqunloadmibs.cmd" rom the \MIBS directory to unload the mibs from NNM.


If the following message displays, you should unload any previous CPQ mibs
and then run cpqloadmibs again.

'[cpqhost.mib:CPQHOST-MIB] - Line 27: MIB_MODULE: "CPQHOST-MIB" loaded from MIB: "CPQHOST.MIB"
Warning:We recommend to resolve duplicate MIB_MODULE names by unloading required MIBs. This 
may create problems later.'

-----

Documentation:

The PDF users guide is located in the \docs directory.

-----

Cluster Discovery:

Only Windows NT Enterprise clusters are identified with a unique cluster
icon.  Other Windows cluster servers will be discovered with the standard
ProLiant icons.

-----

ProLiant Blade discovery:

Blades are discovered as either e-Class or p-Class and displayed on the NNM map.
The operating system on the blades is not discovered and displayed.  

For e-Class blades, the Integrated Administrator (enclosure manager) is also
discovered.

Once the blade systems are discovered and identified, they can be placed into
groups by the user for easier access and management.

-----

Launch to the web-enabled management agents:

You may now right-click on the segment map and select Insight Web Agents on 
the pop-up menu to browse directly to the web-enabled agents.  You can 
also drill down on a node and double-click the web-enabled icon to launch
to the web agents. 

The Insight Web Agents menu item has been added for every system that has 
isCpqInsight set as True.  This does not check to see if the web-enabled agents
are installed or running, so the menu entry may be included on systems without
the web agents.

On the SANworks Management Appliance, the Insight Web Agents link will browse
to the start page for the SANworks Open SAN Manager.  From here, the user can
login and then browse to the HP Management Agents running on the appliance.

-----

In-context launch to HP Systems Insight Manager:

This feature uses the OVwSelection name to browse to a specific device in Systems
Insight Manager.  

This feature depends on the environment variable CIMXEIp.  This environment
variable should be set to the IP address of the Systems Insight Manager server.

If the menu items are launching to an incorrect address, check the CIMXEIp environment
variable.  First, verify the correct address is set in NNM.  Select the root map and
click Map -> Properties.  Click the Applications Tab in the new window and select
Insight Integration for HP OpenView NNM.  Click Configure for this Map and verify 
the correct address is entered in the following window.

Alternatively, you can use the set command to change the address (set CIMXEIp=newaddress).
If the change does not appear to take effect, reboot the system.

-----

RILOE/ILO Discovery and Launch:

RILOE and ILO are now discovered in each server.  When you right-click on a
server on the segment map, if RILOE or ILO is present the Integrated Lights
Out menu option will appear.  This will launch to the RILOE/ILO interface.

Two new fields have been added to the database for RILOE/ILO discovery:
     isCpqRILOE - True/False
     cpqRILOEip - IP Address of the RILOE or ILO

The cpqdis process discovers the RILOE/ILO agent and then retrieves the IP Address
for the cpqRILOEip field.

If the ip address is not appearing correctly, verify the ip address returned by the host
system using the SNMPGET command (snmpget HOSTNAME 1.3.6.1.4.1.232.9.2.5.1.1.5.2).

For example:

D:\snmpget spiral 1.3.6.1.4.1.232.9.2.5.1.1.5.2
232.9.2.5.1.1.5.2 : IpAddress:  172.25.162.159

Also, cpqriloe.exe has been created for launching to the address of the RILOE/ILO
from the server selected on the NNM map.

To display all systems that have been discovered with RILOE/ILO, you can 
perform a find on the isCpqRILOE attribute.  Select Edit->Find->Object by
Attribute.  Scroll down and highlight isCpqRILOE then click Apply.

-----

Insight Summary Launch (for desktops):

The Insight Summary option has been added to the popup menu for desktops to
provide quick access to some of the information provided by the management 
agents.  This option is not intended to provide all the information from 
the management agents.  Full information is available through the web agents.

Any fields that do not apply or could not be determined will appear as blank.

Insight Summary provides the following information:

Product Name
Serial Number
Asset Tag - user configurable field
ROM Version
Physical Memory - total memory installed (Kb)
Operating System - the OS name
Operating System Version - the OS version
Description - the OS description
Processor - only 1 processor (the first) will be displayed
Processor Speed
IDE Drive Model - only one drive (the first) will be displayed
IDE Drive Serial Number
Monitor
Monitor Serial Number
Standard Equipment Status - status reported by the standard equipment MIB
System Information Status - status reported by the system information MIB
Host OS Status - status reported by the host OS MIB
Health Status - status reported by the health MIB
IDE Status - status reported by the IDE MIB


-----

Insight Information Launch (for servers):

The Insight Information option has been added to the popup menu to provide quick
access to some of the information provided by the management agents.  This option
is not intended to provide all the information from the management agents.  That
is available through the web agents.

The MIBs should be loaded to provide the most useful information.  For example,
if the MIBs are not loaded into NNM, then instead of receiving status of other,
ok, degraded, or failed, you will receive status of 0, 1, 2, or 3.  The MIBs 
are now loaded with the integration by default.

If the requested option is not supported by the selected system, you will receive
a message similar to the following: "Note: requested information is not 
available from hostname".  For example, this would occur if you selected the 
IDA Drives option for a system without a drive array.

Summary:
Provides basic information about the system including the status of each
hardware sub-system.

Product Name
Serial Number
Asset Tag - user configurable field
ROM Version
Physical Memory - total memory installed (Kb)
Operating System - the OS name
Operating System Version - the OS version
Description - the OS description
System Role - user configurable field
Standard Equipment Status
System Information Status
Health Status
Host OS Status
Network Interface Status
Threshold Mgmt Status
Drive Array Status
SCSI Status
IDE Status
Storage System Status
Fibre Array Status
Lights-Out Status
Cluster Status
Rack Status


Integrated Management Log:
Displays the information contained in the Integrated Management Log on
the server.

Entries are sorted from old to new, so the most recent entries are displayed
last.  The output window follows the data to display the most recent entries.

Software Versions:
Displays the software version information available to the management agents.
This includes agent and driver information.

System - Processors:

System - Memory:

System - PCI Slots:

Health - Fault Tolerant Fans

Health - Temperatures
Displays information from the health agent temperature sensor table.  Note:
If this value cannot be determined by software, then a value of -1 will be
returned.

Health - Thresholds:
Displays the threshold table for the selected system.  This will display
thresholds set by the management agents.

The following threshold types may be displayed in the table:
 - other(1)   Temperature threshold type could not be determined.
 - blowout(5)
              If a blowout(5) temperature sensor reaches its threshold,
              the fan or fans in the area of the temperature sensor will
              increase in speed in an attempt to reduce the temperature
              before a caution or critical threshold is reached.
 - caution(9)
              If a caution(9) temperature sensor reaches its threshold,
              the cpqHeTemperatureCondition will be set to degraded(3)
              and the system will either continue or shutdown depending
              on the setting of cpqHeThermalDegradedAction.
 - critical(15)
              If a critical(15) temperature sensor reaches its threshold,
              the cpqHeTemperatureCondition will be set to failed(4)
              and the system will shutdown.

Drives - SCSI:
Provides basic information about the SCSI physical drives in the server.

Drives - IDA:
Provides basic information about the Drive Array physical drives in the server.

Drives - Fibre:

Lights-Out Information:

Lights-Out Log:
Displays the information contained in the Remote Insight\Integrated Lights-
Out Log.  This option only appears on systems where the board has been 
discovered.

Entries are sorted from new to old for RILOE and RILOE II, so the most recent
entries are displayed first.

Entries are sorted from old to new for ILO, so the most recent entries are
displayed last.


-----

Graphing Launch:

The Insight Graphs option has been added to the popup menu to provide quick access 
to some of the MIB items provided by the management agents.  The xnmgraph command
is used to present this data.

Percent Disk Space Used
Physical Memory Free
Paging Memory Free
System Temperatures
CPU Utilization - 1 minute
CPU Utilization - 5 minutes
CPU Utilization - 30 minutes
CPU Utilization - 1 hour

Some data items may not be supported on various systems.  For example, the System
Temperatures graph will return -1 for values that are not supported.  These will
appear as N/A in the web-enabled agents.


More useful information will be provided for data collection if the mibs are 
loaded into NNM.  For example, instead of getting a label of 3 for data collection,
you would get cpqHoCpuUtilFiveMin as a label with the mibs loaded.

Note: When using the integration, you should not load the traps from the MIBs.
The trap definitions are provided in the configuration files instead.


The Insight Graphs option appears on all items with the isCpqInsight property
set to true.  In some cases, the graphs will not apply to the system where 
the option is displayed.  For example, the ProLiant BL10e Enclosure will have
this options, but the graphs will return no data.


-----

Integration with HP OpenViews Java-based web interface:

Icons are now provided for the NNM web interface.  With these icons, systems
will be correctly identified when the NNM web interface is used.

In addition to the icons, a registration file is provided to allow access
to Systems Insight Manager and the Insight Management Agents from the NNM web
interface.

The file \imovwin31\www\insight contains the registration information for
these applications.  In order to launch Systems Insight Manager in context, the
file must be edited to include the IP address of the users Insight Manager
7 server.

1. Open the file "insight".
2. Search for IM7IPADDRESS.
3. Replace with the name or ip address of your server.
4. Save the file.

The file "insight" is copied to the %OV_WWW_REG%\jovw\%LANG% directory.



A registration file is also provided to allow access to the Systems 
Insight Manager homepage from the launcher tools menu.

The file \imovwin31\www\insighthome contains the registration information
for this application.  In order to launch to Systems Insight Manager, the
file must be edited to include the IP address of the users Systems Insight
Manager server.

1. Open the file "insighthome".
2. Search for IM7IPADDRESS.
3. Replace with the name or ip address of your server.
4. Save the file.

The file "insighthome" is copied to the %OV_WWW_REG%\launcher\%LANG% directory.


-----


New Trap Information:


The latest HP alarms send several new pieces of information.  The
HP trap definition files have been updated to take advantage of
this.

For example, the following is an alarm from a fault tolerant power 
supply:

- Major Tue Nov 06 11:07:49  CRONUS.mse.com    HP - Fault Tolerant 
Power supply Failed. (Chassis: 0, Bay: 2, Model: 169282-002, Serial#:
51262E2+G624Q, Firmware: 6.31, Spare#: )

If one or more of the fields in the alarm is blank (the spare number 
in the example above), the most likely explanation is that information
was not sent in the alarm.

In the power supply ok alarm shown below, only the chassis number and 
bay number are sent in the alarm.  The other fields are blank.

- Normal Tue Nov 06 11:02:33  CRONUS.mse.com   HP - Fault Tolerant 
Power supply OK. (Chassis 0, Bay 2, Model , Serial# , Firmware , Spare# )


Also, all the HP alarm message have been update to begin with "HP -".  

If you want to see all the HP alarms that have been received, in the
All Alarms Browser select View - Set Filters.  In this window, select
the Match Message String.  Then enter "HP -" in the string field.

This will display all alarms containing the string "HP -".

---------------------------------------------------------------------


Advanced Configuration:

-----

Setting up the integration with an NNM Remote Console:


NNM Remote Console installations are not supported by the integration
module.  However, the following steps can be followed in order have
the integration function on a Remote Console install.

This information only applies if the management server and remote console
are both Windows systems.

1. Follow the installation procedure to install the integration on
   the NNM server.

2. Copy the %OV_BIN%\cpq directory from the NNM server to the NNM
   Remote Console %OV_BIN% directory so that both systems have the
   %OV_BIN%\cpq directory and all the contained files.

3. Set the CIMXEip environment variable on the Remote Console 
   system to the IP address of the Systems Insight Manager server in 
   order to utilize the launch to Systems Insight Manager.  

4. Restart the system once the environment variable has been
   created and set.


Again, this configuration is not supported and not been extensively
tested.




---------------------------------------------------------------------

Troubleshooting:

Several tools are provided in the %OV_MAIN_PATH%\install\tools\cpq 
directory for validating the installation.

Refer to the Insight Integration for HP OpenView NNM Users Guide 
PDF for more information on troubleshooting.  

-----


Undefined Alarms in the event log:

If alarms similar to the following continue to appear in the alarm browser, the trapd
definitions for the integration module may not be installed correctly.  Verify that the
trap definition files are properly installed.  

Reload the trap definitions by executing "xnmevents -load newtraps.cpq", then stop and
restart the OpenView services (ovstop and ovstart).

In the messages below, "nnmsystem.mse.com" is the name of the server running Network Node Manager.

- Normal Fri Sep 13 10:30:16  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.2.0.3001 (enterprise:.1.3.6.1.4.1.232.1.3.2 generic:6 specific:3001), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.2 (Integer): 7215
- Normal Fri Sep 13 10:30:16  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.11.0.3003 (enterprise:.1.3.6.1.4.1.232.1.3.11 generic:6 specific:3003), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.11 (Integer): 7215
- Normal Fri Sep 13 10:30:16  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.3.0.3002 (enterprise:.1.3.6.1.4.1.232.1.3.3 generic:6 specific:3002), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.3 (Integer): 7215
- Normal Fri Sep 13 10:30:17  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.2.0.3001 (enterprise:.1.3.6.1.4.1.232.1.3.2 generic:6 specific:3001), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.2 (Integer): 7217
- Normal Fri Sep 13 10:30:17  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.2.0.3001 (enterprise:.1.3.6.1.4.1.232.1.3.2 generic:6 specific:3001), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.2 (Integer): 7220
- Normal Fri Sep 13 10:30:17  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.11.0.3003 (enterprise:.1.3.6.1.4.1.232.1.3.11 generic:6 specific:3003), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.11 (Integer): 7217
- Normal Fri Sep 13 10:30:17  nnmsystem.mse.com        Received event .1.3.6.1.4.1.232.1.3.3.0.3002 (enterprise:.1.3.6.1.4.1.232.1.3.3 generic:6 specific:3002), no format in trapd.conf. 1 args: [1] private.enterprises.compaq.cpqStdEquipment.3.3 (Integer): 7217

-----

Nodes not classified as HP servers:

There are several possible causes for this situation.  First you should verify
communication with the system.

Right click on the node and select Object Properties.

Open a command prompt and try to ping the selection name displayed in the
Object Properties dialog.

If you cannot ping the selection name, add it to the hosts file on the
system.  Then the device should be classified correctly.


If this has no effect, then verify the console can communicate 
with the HP Management Agents.

Run cpqchksnmp in the %OV_MAIN_PATH%/install/tools/cpq directory
to verify if a system is running SNMP.

Example: cpqchksnmp 172.25.162.30 public

-

Run cpqchkagnt in the %OV_MAIN_PATH%/install/tools/cpq directory
to verify if a system is running the Compaq agents.

Example: cpqchkagnt -n 172.25.162.30

-

Perform the following steps to verify that OpenView NNM can 
get the operating system information from the desired system.

1. Run the OpenView MIB browser %OV_BIN%/xnmbrowser.
2. Enter the hostname or IP address of the system.
3. Enter the community name to use for communication.
4. Enter .1.3.6.1.4.1.232.11.2.2 for the MIB Object ID.
5. Click Start Query.

The results should display the operating system name and version,
similar to the following (depending on the operating system)
in the MIB Values field:

1.0:Microsoft Windows NT Server
2.0:4.00
3.0:Service Pack 6 1381 Uniprocessor Free
4.0:3
5.0:3

If you cannot obtain any results, then the system cannot
communicate with the HP Management Agents.



-----

Insight Inforamtion options return errors when querying a node:


SNMP must be installed and configured for the Insight Information menu
items to work properly.  In addition, the correct community string must
be configured within NNM.


"Could not perform SNMP get from HOSTNAME: No response arrived before timeout"

Verify that the correct SNMP community string being used.



"Note: requested information is not available from HOSTNAME"

In most cases, this message is not an error.  The most likely cause of this 
message is requesting information that is not implemented on the target node.
For example, selecting Insight Information -> Drives - SCSI on a system that
only has a drive array (no SCSI devices) will result in this message.



---------------------------------------------------------------------

Known Issues:


The cpqmap process may exit ungracefully when closing the NNM map.
The following is an example of the Error Alarm.

A Major  Fri 09 03 15:12:49  nnmsystem   pmd lost contact with OVsnmpAPI_application: pid 2404 ungracefully disconnected from pmd with 0 pending events: sendto() failed: Software caused connection abort (390 byte send() failed: Software caused connection abort (errno 231: ERROR 231))

-----

